JP4048481B2 - Electron beam proximity exposure method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an electron beam proximity exposure method, and more particularly to an electron beam proximity exposure method in which a mask pattern of a mask arranged in proximity to a semiconductor wafer is transferred to a resist layer on the wafer at an equal magnification using an electron beam.
[0002]
[Prior art]
Conventionally, this type of electron beam proximity exposure apparatus is disclosed in US Pat. No. 5,831,272 (corresponding to Japanese Patent No. 2951947) (Patent Document 1).
[0003]
FIG. 16 is a view showing a basic configuration of the electron beam proximity exposure apparatus. The electron beam proximity exposure apparatus 10 includes an electron beam source 14 that mainly generates an electron beam 15, an electron gun 12 that includes a lens 16 and a shaping aperture 18 that convert the electron beam 15 into a parallel beam, main deflectors 22, 24, and The scanning unit 20 includes sub-deflectors 26 and 28 and scans an electron beam parallel to the optical axis, and a mask 30.
[0004]
The mask 30 is disposed so as to be close to the wafer 40 having the resist layer 42 formed on the surface (for example, the gap is 50 μm). In this state, when the electron beam is irradiated perpendicularly to the mask 30, the electron beam that has passed through the mask pattern of the mask 30 is irradiated to the resist layer 42 on the wafer 40.
[0005]
Further, the scanning means 20 controls the deflection of the electron beam so that the electron beam 15 scans the entire surface of the mask 30 as shown in FIG. As a result, the mask pattern of the mask 30 is transferred to the resist layer 42 on the wafer 40 at the same magnification.
[0006]
The electron beam proximity exposure apparatus 10 is provided in a vacuum chamber 50 as shown in FIG. Further, in the vacuum chamber 50, an electrostatic chuck 60 for attracting the wafer 40 and the wafer 40 attracted to the electrostatic chuck 60 are moved in two horizontal orthogonal directions and rotated in a horizontal plane. A θXY stage 70 is provided. The θXY stage 70 moves the wafer 40 by a predetermined amount each time the mask pattern is transferred at an equal magnification, thereby allowing a plurality of mask patterns to be transferred onto one wafer 40. Such a transfer method is referred to as a die-by-die method. In FIG. 18, in order to conduct the wafer 40, conduction pins 80 pressed against the upper surface of the wafer 40 are provided.
[0007]
By the way, the wafer is exposed a plurality of times using a plurality of masks each having a different mask pattern, thereby forming an integrated circuit. When exposing each mask pattern, it is necessary to relatively position the mask and the wafer so that the mask pattern to be exposed has a predetermined positional relationship with the already exposed mask pattern.
[0008]
However, even if the positioning of the mask and the wafer is performed with high accuracy, for example, when the wafer expands and contracts with respect to the design value through an exposure process or the like, there is a problem that the exposed mask patterns are shifted from each other.
[0009]
In order to solve this problem, a magnification correction amount detection method for obtaining a magnification correction amount for a mask and a wafer has been proposed (Patent Document 2). This document discloses a method in which a mask is placed close to a wafer and a mask pattern is exposed to X-rays on the wafer, and the magnification correction is performed by locally heating the mask to thermally deform it, It is disclosed that the mask is deformed by applying stress from the outside.
[0010]
[Patent Document 1]
US Pat. No. 5,831,272
[0011]
[Patent Document 2]
JP 2000-353647 A
[0012]
[Problems to be solved by the invention]
However, the conventional die-by-die method takes time required for positioning (alignment) between the mask and the wafer, and as a result, has a big problem that the throughput is low.
[0013]
On the other hand, when the alignment time is shortened, there is a great concern that the positioning accuracy is sacrificed. Further, for example, when the wafer is expanded or contracted with respect to the design value through an exposure process or the like, even if the mask is deformed by heat or external force and the magnification between the mask and the wafer is corrected as described in Patent Document 2, There is also a problem that it is difficult to deform with high accuracy, and it is necessary to check the degree of deformation of the mask, and the magnification cannot be easily corrected.
[0014]
The present invention has been made in view of such circumstances, and provides an electron beam proximity exposure method capable of reducing the time required for alignment as much as possible and transferring a mask pattern to each die of a wafer easily and accurately. For the purpose.
[0015]
  In order to achieve the above object, the present invention provides a mask pattern formed on a mask by arranging the mask close to the wafer, performing alignment for each die of the wafer, and scanning the mask with an electron beam. In the electron beam proximity exposure method for transferring to the resist layer on the wafer, step 1 which is a step of storing the alignment information of each die in a predetermined process, and alignment of the die transferred next from the stored alignment information Step 2 which is a step of calculating a position, Step 3 which is a step of positioning a die to be transferred next based on the calculation result, and a position where the die is transferred to the transfer position. A first mark for alignment provided on the mask and an alignment provided for each die Step 4 which is a step of imaging the second mark and calculating the positioning result of the die from the captured information, the alignment position calculated in Step 2 and the positioning result calculated in Step 4 Step 5 which is a comparison step, and a step of performing alignment so that the amount of positional deviation between the mask and the die is less than a predetermined value β based on the comparison result and / or controlling the incident angle of the electron beam And Step 7 which is a step of transferring the mask pattern to a resist layer on the wafer.When the comparison result in step 5 is less than the predetermined threshold value α, the frequency that is the comparison result is counted, and when the count result is the predetermined frequency M, the frequency is determined in the subsequent steps. Step 4, Step 5 and Step 6 are omitted for every number of dies.An electron beam proximity exposure method is provided.
[0016]
According to the present invention, in the die-by-die type electron beam proximity exposure method, the calculated alignment position of the next transferred die is compared with the positioning result calculated from the imaged information. Is equal to or greater than a predetermined value β, alignment is performed so that the comparison result is less than the predetermined value β and / or the incident angle of the electron beam is controlled. On the other hand, when the comparison result is less than the predetermined value β, the transfer is performed as it is. Therefore, the time required for alignment can be shortened as much as possible, and the mask pattern can be easily and accurately transferred to each die of the wafer.
[0017]
“During a predetermined process” means when the alignment of a die is completed and accurate alignment information of the die is obtained, or when the transfer of the die is completed after the alignment is completed. To do.
[0018]
Further, “controlling the incident angle of the electron beam” in step 6 means that the electron beam proximity exposure apparatus is set so that the incident angle of the electron beam becomes a desired value in the transfer in step 7 which is the next process. This refers to preparing control data for the sub deflector.
[0019]
In the die-by-die type electron beam proximity exposure method, “alignment” refers to the relative relationship between a mask (first mark for alignment in the present invention) and a wafer (second mark for alignment in the present invention). This refers to an operation for obtaining an accurate misregistration amount and performing mechanical alignment so that the misregistration amount is less than a predetermined value, controlling an incident angle of an electron beam, or both.
[0020]
  In the present inventionIn particular,When the comparison result in step 5 is less than the predetermined threshold value α, the frequency that becomes the comparison result is counted, and when the count result reaches the predetermined frequency M, every predetermined number in the subsequent steps. Step 4, Step 5 and Step 6 are omitted in the dieI am doing so.
[0021]
Thus, if the error in alignment is less than the threshold value α, the feedback type alignment is not performed, and there is a high probability that the predetermined accuracy is obtained even in the feedforward type positioning. Therefore, the feedback type alignment is not performed for each die, for example, the feedback type alignment is performed for every two dies and every three dies, and the feed forward type positioning can be performed for the other dies. Thereby, the time required for alignment can be significantly shortened.
[0022]
In the present invention, the transfer for each die of the wafer is preferably performed in the order from the outer peripheral die of the wafer toward the central die of the wafer.
[0023]
When a mask or a wafer expands or contracts with respect to a design value through an exposure process or the like, an error at the outer peripheral portion of the wafer is usually larger than an error at the central portion of the wafer. Therefore, when the transfer for each die of the wafer is performed in this order from the die on the outer peripheral portion of the wafer toward the die on the central portion of the wafer, an error in alignment with the die on the outer peripheral portion of the wafer is a threshold α. If it is less than this, there is a high probability that the error in the alignment of the die at the center of the wafer is smaller than this. Therefore, it is also possible to omit all the feedback type alignment in the die after the error in alignment is less than the threshold value α. Thereby, the time required for alignment can be significantly shortened.
[0024]
Further, in the present invention, the comparison results of the step 5 in one wafer are totaled, statistical processing is performed, and the frequency of the die that omits the steps 4, 5 and 6 in subsequent wafers is determined based on the processing result. It is preferable to increase or decrease.
[0025]
There is a strong tendency that the mechanical characteristics, thermal characteristics, etc. are uniform between wafers in the same lot. Therefore, if the results of processing one wafer of the same lot are statistically processed, the results for other wafers can often be inferred. In such a case, even if the frequency of the die for performing the feedback alignment on the first wafer is slightly increased, the feedback alignment is performed on the second wafer, the third wafer, etc. It is often possible to obtain statistical results that are sufficient to reduce the frequency of dies performed. Therefore, by adopting this method, the time required for alignment can be greatly shortened.
[0026]
In the present invention, the alignment in step 6 and / or the control of the incident angle of the electron beam may be performed by the amount of deviation in the X direction between the first mark and the second mark, the first mark and the second mark. It is preferable to make the deviation based on the amount of deviation in the Y direction from the second mark, the amount of deviation in the θ direction between the first mark and the second mark, and the mask distortion of the mask.
[0027]
This is because positioning accuracy can be easily obtained by performing alignment and / or controlling the incident angle of the electron beam based on the above four values.
[0028]
Further, in the present invention, the alignment in the step 6 includes the amount of deviation in the X direction between the first mark and the second mark, and the Y direction between the first mark and the second mark. Preferably, the electron beam incident angle is controlled based on the shift amount in the θ direction between the first mark and the second mark and the mask distortion of the mask.
[0029]
  As described above, the correction of the deviation amount between the marks in the XY directions can be easily performed by fine movement adjustment in the XY directions. Further, the correction of the shift amount in the θ direction and the correction of the mask distortion amount of the mask can be easily performed by controlling the incident angle of the electron beam.
  In particular, when correction of the shift amount in the θ direction is performed by fine adjustment of the wafer stage or the like, relative rotation between the mask and the wafer occurs, and further correction of the shift amount in the θ direction is necessary for dies at different positions on the wafer. And the calculation becomes complicated. On the other hand, if the shift amount correction in the θ direction is performed by controlling the incident angle of the electron beam, the relative rotation between the mask and the wafer does not occur, so that such a problem does not occur.
[0030]
  In the present invention, it is preferable that a plurality of the second marks are provided around the dies, and the first marks are provided corresponding to the second marks.
  In the present invention, it is preferable that the second mark is provided on each side of each die having a rectangular shape.
  Furthermore, in the present invention, it is preferable that a plurality of the first marks and the second marks that are provided are picked up by an image pickup device that is provided corresponding to the first marks and the second marks.
[0031]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, preferred embodiments of the electron beam proximity exposure method according to the present invention will be described. First, a preferred embodiment of an electron beam proximity exposure apparatus (mainly hardware) used in the electron beam proximity exposure method according to the present invention will be described with reference to the accompanying drawings, and then an electron beam proximity exposure method using the same. The flow (mainly software) will be described.
[0032]
FIG. 1 is a top view of a transfer portion of an electron beam proximity exposure apparatus used in the electron beam proximity exposure method according to the present invention. FIG. 2 is a view taken along line AA in FIG.
[0033]
As shown in these drawings, the electron beam proximity exposure apparatus is provided with four microscope imaging devices AX1, AY1, AX2, and AY2 facing the mask 32. These microscope imaging devices AX1, AY1, AX2, and AY2 are arranged so that the photographing optical axis is inclined with respect to the mask surface so as not to block the electron beam during exposure with the electron beam. The main configuration of the electron beam proximity exposure apparatus is the same as that shown in FIGS. 16 to 18, and thus detailed description thereof is omitted.
[0034]
FIG. 3 is an enlarged top view of the transfer portion of the electron beam proximity exposure apparatus. In the figure, a wafer 44 is attracted by an electrostatic chuck 60 on a θXY stage 70.
[0035]
The wafer 44 includes four sets of wafer marks M for positioning each chip of the wafer 44 and measuring the expansion / contraction rate of each chip._WX1, M_WX2, M_WY1, M_WY2(See FIG. 4) is provided outside the region where the chip is formed.
[0036]
On the other hand, the mask 32 has a mask pattern formed in the area indicated by the broken line, and the wafer mark M is outside the area indicated by the broken line._WX1, M_WX2, M_WY1, M_WY2The four mask marks M for detecting the displacement of the mask 32 and the wafer 44 in the x direction and the y direction, the displacement of the rotation direction of the xy plane, and the expansion / contraction rate of the chip in the x direction and the y direction._MX1, M_MX2, M_MY1, M_MY2Is provided.
[0037]
Details of the wafer mark and the mask mark will be described with reference to FIGS. 5 is an enlarged view of a portion indicated by reference numeral B in FIG. 3, and FIG. 6 is a cross-sectional view taken along the line CC in FIG. As shown in FIG. 5, the mask 32 has a mask mark M._MX1, And a wafer mark M below the mask 32 through the mask 32._WX, M_WXCan be seen through.
[0038]
Mask mark M formed on mask 32_MX1Is composed of 5 × 3 small openings, while the wafer mark M_WXIs composed of 14 × 3 convex portions (see FIGS. 5 and 6). The positioning marks formed on the wafer or mask are not limited to this embodiment.
[0039]
Next, the microscope imaging apparatus will be described. Here, the microscope imaging apparatus AX1 will be described, but the other microscope imaging apparatuses AY1, AX2, and AY2 have the same configuration. FIG. 7 shows an outline of the microscope imaging apparatus AX1. As shown in the figure, the microscope imaging apparatus AX1 is arranged such that the incident angle of the optical axis is a predetermined angle φ with respect to the mask 32 arranged horizontally.
[0040]
A cover glass 91 is attached to the front surface of the microscope objective lens 90. A conductive thin film 91A is deposited on the surface of the cover glass 91, and the conductive thin film 91A is grounded via the conductive holding member 92 and the housing of the microscope imaging apparatus AX1. This prevents electrons or secondary electrons scattered during transfer by the electron beam from being charged on the surface of the cover glass 91.
[0041]
As the conductive thin film 91A, a tin oxide film or an indium tin oxide film (ITO) is used. In this embodiment, the conductive thin film 91A is vapor-deposited on the surface of the cover glass 91. However, in the case of a microscope imaging apparatus in which the cover glass is not provided, the surface of the microscope objective lens 90 is electrically conductive. A conductive thin film is deposited so that the conductive thin film is grounded.
[0042]
In addition, a mechanical shutter mechanism may be provided instead of the conductive thin film 91A, and the shutter mechanism may be closed to shield the optical member of the microscope image pickup device during the transfer by the electron beam, and the shutter mechanism may be opened during the image pickup. . In this case, a shutter mechanism that does not charge electrons is used.
[0043]
Illumination means is provided inside the microscope imaging apparatus AX1. That is, the illumination means is composed of a white light source 93, a lens 94, a reflection mirror 95 and a half mirror 96, and the white illumination light emitted from the white light source 93 is made into substantially parallel light by the lens 94 and is reflected by the reflection mirror 95. The mask 32 and the wafer 44 are illuminated through the half mirror 96, the objective lens 90, and the cover glass 91.
[0044]
The scattered light from the mask mark of the mask 32 and the wafer mark of the wafer 44 thus illuminated enters the imaging unit 97 via the microscope objective lens 90 and the half mirror 96 and is imaged.
[0045]
As shown in FIG. 8A, the mask mark M_MXAnd wafer mark M_WXThe length of each mark and the incident angle φ of the photographic optical axis are determined so that a part of the mark is positioned on the focal plane F of the microscope imaging apparatus. In FIG. 8A, P is the specular reflected light of white illumination light, and Q and R are mask marks M, respectively._MXAnd wafer mark M_WXThe scattered light of white illumination light at G, G is the distance between the mask 32 and the wafer 44.
[0046]
FIG. 8B is an image showing a state in which the scattered lights Q and R of the white illumination light are imaged by the microscope imaging device.
[0047]
Next, a method for detecting the deviation between the mask and the wafer based on the mask mark and wafer mark imaged as described above will be described. As shown in FIG. 9A, a focused portion (a portion surrounded by a frame) is continuously extracted in the x direction from the captured mask mark and wafer mark images. FIG. 9B shows the luminance level at each position in the x direction of the image extracted in this way.
[0048]
Here, as shown in FIG. 9B, two mask marks M_MX, And one wafer mark M_WXFor the luminance level corresponding to, the center peak position of each of the three peaks is obtained, and the distance x between the peaks is determined.₁, X₂Ask for. And mask mark M_MXAnd wafer mark M_WXAnd the positional deviation amount Δx in the x direction is expressed by the following equation:
[0049]
[Expression 1]
Δx = (x₁-X₂) / 2
Can be obtained.
[0050]
It should be noted that the two displacement amounts of the mask 32 and the wafer 44 in the x direction can be detected by the microscope imaging devices AX1 and AX2, and the two displacements of the mask 32 and the wafer 44 in the y direction can be detected by the microscope imaging devices AY1 and AY2. The amount can be detected.
[0051]
Next, a method of measuring the expansion / contraction rate in the x direction and y direction of each chip of the wafer will be described. When measuring the expansion / contraction ratio of the chip, the chip to be measured and the mask 32 are positioned. When measuring the expansion / contraction ratio of the chip in the y direction, the three microscope imaging devices AX1, AY1, and AX2 are used. The θXY stage 70 is moved in the x and y directions so that the detected positional deviation amount becomes zero at the same time, and the θXY stage 70 is rotated in the xy plane. In this embodiment, a positioning mark that matches the center of the chip with the center of the mask is formed.
[0052]
Then, a positional deviation amount Δy detected by the remaining microscope imaging device AY2 is obtained. This positional deviation amount Δy is determined by the two wafer marks M_WY2And mask mark M_MY2And each distance with y as shown in FIG.₁, Y₂Then, the following formula:
[0053]
[Expression 2]
Δy = (y₁-Y₂) / 2
Can be expressed as The positional deviation amount Δy obtained in this way is used as the two wafer marks M from the center of the chip._WY2By dividing by the reference length in the y direction with respect to the center of the chip, the expansion / contraction rate of the chip in the y direction can be obtained.
[0054]
Similarly, when measuring the expansion / contraction ratio of the chip in the x direction, the θXY stage 70 is moved in the x direction so that the amount of displacement detected by the three microscope imaging devices AX1, AY1, and AY2 becomes zero simultaneously. While moving in the y direction, the θXY stage 70 is rotated in the xy plane, and a positional deviation amount Δx detected by the remaining microscope imaging device AX2 is obtained. This positional deviation amount Δx is determined by the two wafer marks M_WX2And mask mark M_MX2For each distance x as shown in FIG.₁, X₂Then, the following formula:
[0055]
[Equation 3]
Δx = (x₁-X₂) / 2
Can be expressed as The positional deviation amount Δx obtained in this way is determined from the two wafer marks M from the center of the chip._WX2By dividing by the reference length in the x direction with respect to the center of the chip, the expansion / contraction rate in the x direction of the chip can be obtained.
[0056]
FIG. 10 is a block diagram showing an embodiment of the controller of the electron beam proximity exposure apparatus according to the present invention. In the figure, a central processing unit (CPU) 100 performs overall control of the entire apparatus. As described above, the processing for obtaining the expansion and contraction rates in the x and y directions of the wafer, wafer positioning control, and exposure time are performed. Performs deflection control of the electron beam.
[0057]
Each image signal obtained by imaging with the four microscope imaging devices AX1, AY1, AX2, and AY2 is applied to the signal processing circuit 102. The signal processing circuit 102 calculates four misregistration amounts between the mask mark and the wafer mark based on each input image signal.
[0058]
The CPU 100 moves the θXY stage 70 in the x and y directions via the stage driving circuit 104 so that the four positional shift amounts input from the signal processing circuit 102 become zero, and moves the θXY stage 70 in the xy plane. The wafer is rotated, thereby positioning the wafer with high precision (fine alignment).
[0059]
The CPU 100 supplies correction data corresponding to the expansion / contraction ratio of the wafer together with the deflection amount data for scanning the mask to the digital arithmetic circuit 106, and the digital arithmetic circuit 106 performs a digital signal for scanning the mask based on the deflection amount data. Is output to the main DAC / AMP 108, and the transfer magnification in the x and y directions is changed in proportion to the expansion and contraction rates in the x and y directions of the wafer based on the correction data, and the mask distortion is corrected as will be described later. The digital signal for this is output to the sub DAC / AMP 110.
[0060]
The main DAC / AMP 108 converts the input digital signal into an analog signal, amplifies it, and outputs it to the main deflectors 22 and 24 shown in FIG. As a result, the electron beam 15 is deflected so as to scan the entire surface of the mask 30 as shown in FIG. 17 while maintaining a state parallel to the optical axis.
[0061]
The sub DAC / AMP 110 converts the input digital signal into an analog signal, amplifies it, and outputs it to the sub deflectors 26 and 28 shown in FIG. Thereby, the incident angle of the electron beam 15 on the mask 32 is controlled as shown in FIG.
[0062]
In FIG. 10, the laser interferometer L_X, L_YIs for monitoring the amount of movement of the θXY stage 70.
[0063]
As shown in FIG. 11, when the incident angle of the electron beam 15 to the mask 32 is Ψ and the distance between the mask 32 and the wafer 44 is G, the shift amount δ of the transfer position of the mask pattern by the incident angle Ψ is The following formula,
[0064]
[Expression 4]
δ = G · tan Ψ
It is represented by In FIG. 11, the mask pattern is transferred to a position shifted from the normal position by the shift amount δ.
[0065]
Therefore, the transfer magnification can be changed by changing the incident angle Ψ according to the scanning position of the electron beam. The incident angle Ψ is set to 0 at the mask center, and increases as the distance from the mask center increases.
[0066]
FIG. 12 is a flowchart showing the operation procedure of the electron beam proximity exposure method according to the present invention. As shown in the figure, a mask is first loaded on the electron beam proximity exposure apparatus (step S10), and then the wafer is positioned on the electrostatic chuck 60 on the θXY stage 70, and then attracted and fixed by the electrostatic chuck 60. As a result, the wafer is loaded (step SS12). Subsequently, the wafer is brought into conduction with conduction pins or the like so that electrons are not charged on the wafer (step S14).
[0067]
Next, the height is detected by the z sensor for detecting the height of the wafer, the wafer height is adjusted (step S16), and then the wafer is roughly aligned (course alignment) (step S18). . Subsequently, the wafer is moved to the transfer position (step S20).
[0068]
Next, the gap (GAP) between the mask and the wafer is adjusted (step S22). As shown in FIG. 8A, the gap G between the mask and the wafer is such that the imaging surface of the microscope imaging device and the mask mark M_MXIntersection point Q and wafer mark M_wxA line segment QR with the intersection point R is obtained based on the photographed image, and from this line segment QR and the incident angle φ of the photographing optical axis,
[0069]
[Equation 5]
G = QR ・ sin φ
Can be obtained. Details are disclosed in JP-A No. 2000-356511. Note that the method of measuring the gap G is not limited to this embodiment.
[0070]
The interval G is adjusted so that the interval G measured as described above becomes a predetermined value (for example, 50 μm). The value of the interval G (interval value) is added to the correction arithmetic circuit 106A. The correction arithmetic circuit 106A corresponds to a circuit that performs deflection control of the sub deflectors 26 and 28 in the digital arithmetic circuit 106 shown in FIG.
[0071]
Next, as described in FIG. 3, the expansion / contraction rate is measured for each chip (MAG measurement (chip)) (step S23).
[0072]
Thereafter, the mask and the chip on the wafer are accurately positioned (fine alignment) using the four microscope imaging devices AX1, AY1, AX2, and AY2 (step S24), and then the mask pattern is transferred to the wafer by the electron beam. (Step S26).
[0073]
At the time of this transfer, the correction arithmetic circuit 106A makes the wafer expansion / contraction rate ε_x, Ε_YThe transfer magnification is changed based on the above, and the incident angle of the electron beam to the mask pattern is controlled (tilt correction) so as to correct the mask distortion. Note that data indicating the mask distortion measured in advance in step S32 is input to the correction arithmetic circuit 106A, and the correction arithmetic circuit 106A receives the mask distortion as shown in FIG. 13A, for example. Further, the tilt correction of the electron beam is performed so that the mask pattern without the mask distortion as shown in FIG. 13B is transferred.
[0074]
When the transfer of the mask pattern to the wafer is completed, the wafer is unloaded (step S28).
[0075]
Next, the flow (mainly software) of the electron beam proximity exposure method using the electron beam proximity exposure apparatus will be described. FIG. 14 is a diagram showing the arrangement of the dies D, D... On the wafer 44. The wafer 44 has a diameter of 300 mm, and is configured such that 88 square dies D each having a side of 25 mm are arranged.
[0076]
The illustrated example shows a state in which the transfer starts from the outermost die D1 and proceeds in the clockwise direction to the die D2 and the die D3, and the transfer of the die D15 is completed. When the transfer further proceeds, the transfer proceeds in a spiral manner in the order from the die D at the outer peripheral portion of the wafer 44 toward the die D at the central portion of the wafer 44 as indicated by the broken line.
[0077]
FIG. 15 is a flowchart for explaining the flow of the electron beam proximity exposure method. Hereinafter, the flow of the electron beam proximity exposure method will be described with reference to FIGS. 14 and 15.
[0078]
The alignment information of the die D15 at a predetermined process (in FIG. 14, when the transfer of the die D15 is completed) is stored (step S102). Then, the alignment position of the die D16 to be transferred next is calculated from the stored alignment information (step S103).
[0079]
Based on the calculation result, the next transfer die D16 is positioned at the transfer position (step S104). Then, a mask mark (M, which is a first mark for alignment provided on the mask 32)._MX, M_MY, M_MX1, M_MX2, M_MY1, M_MY2) And a wafer mark (M) which is a second mark for alignment provided on the die D16._WX, M_WY, M_WX1, M_WX2, M_WY1, M_WY2) And the positioning result of the die D16 is calculated from the captured information (step S105).
[0080]
The alignment position calculated in step S103 is compared with the positioning result calculated in step S105. Then, alignment is performed and / or the incident angle of the electron beam is controlled so that the amount of positional deviation between the mask 32 and the die D16 is less than a predetermined value β based on the comparison result (step S106). As a specific example of the predetermined value β, 5 nm can be adopted. Next, the mask pattern is transferred to the resist layer on the wafer 32 (step S107).
[0081]
Next, it is determined whether or not the positional deviation amount between the mask 32 and the die D16, which is the comparison result in step S106, is less than a predetermined threshold value α (step S108). As a specific example of the predetermined threshold α, 5 nm can be adopted.
[0082]
If the amount of positional deviation is greater than or equal to the predetermined threshold value α, the process returns to step S102, and alignment of the next die (die D17 in this example) is started.
[0083]
When the amount of positional deviation is less than the predetermined threshold value α, the frequency that is less than the threshold value α is counted, and when the count result becomes the predetermined frequency M, in the subsequent steps, the frequency is changed every predetermined number. In the die, step S105 and step S106 (corresponding to step 4, step 5 and step 6 in claim 2) are omitted. Hereinafter, it demonstrates according to a flow.
[0084]
In the die D16, when the positional deviation amount is less than the threshold value α, 1 is added to the number N of times that the positional deviation amount has been less than the threshold value α in the conventional die D. That is, N = N + 1 is calculated (step S109). Then, the value of N is compared with a predetermined frequency M (step S110).
[0085]
If the value of N is less than the predetermined frequency M, the process returns to step S102, and the alignment of the next die (die D17 in this example) is started.
[0086]
In the case where the value of N is a predetermined frequency M or more than M, the steps S105 and S106 are omitted in the subsequent predetermined number of dies (for example, every three sheets).
[0087]
More specifically, in the dies D17, D18 and D19, the alignment information of the die D is stored (step S202), the alignment position of the die D to be transferred next is calculated (step S203), and the next based on the calculation result. Only positioning of the die D to be transferred to the transfer position (step S204) and transfer (step S207) are performed. In this step, Step S202, Step S203, Step S204, and Step S207 have the same processing contents as Step S102, Step S103, Step S104, and Step S107, respectively.
[0088]
Next, it is determined whether or not the number of alignment skips (three in this example) has expired (step S212). If the number of skips has not been completed, the process returns to step S202, and this loop is repeated to complete the number of skips. If so, the process returns to step S102, and alignment of the next die (in this example, die D20) is started.
[0089]
In the case of the flow returning from step S212 to step S102, in this example, the alignment of the die D20 is started. In the processing of the die D20, when the amount of positional deviation is less than the predetermined threshold value α in step S108, The process proceeds to step S109. In step S109, since 1 is further added to the value of N that has already become the predetermined frequency M or more, the process inevitably proceeds to step S202, and dies for every predetermined number thereafter (for example, every three sheets). Then, the process of step S105 and step S106 will be omitted.
[0090]
In the flowchart of FIG. 15, in step S108, it is determined whether or not the value of the comparison result is less than a predetermined threshold value α, and for subsequent dies (for example, three), step S105 and step S106 are performed. However, the predetermined number of dies that skip the process may be changed depending on the value of the comparison result.
[0091]
Further, as described above, when an error occurs due to expansion or contraction of a mask or a wafer with respect to a design value through an exposure process or the like, the error at the outer periphery of the wafer is usually larger than the error at the center of the wafer. Therefore, after the predetermined die D, the predetermined number of dies to be skipped is greatly increased or divided into several stages (the predetermined interval of the skipped dies is widened). For the following dies from the predetermined die D, an algorithm that omits all the steps S105 and S106 can be adopted.
[0092]
Although not described in this embodiment, the comparison results for one wafer are aggregated and statistical processing is performed as described in claim 4, and statistical processing is performed. Based on the processing results, steps S 105 and S 106 for subsequent wafers are performed. A method of increasing or decreasing the frequency of the die that omits this step can also be adopted.
[0093]
For example, in the first wafer of the same lot, the ratio of dies that omit the steps S105 and S106 in all dies was 20%. However, depending on the result of statistical processing on the first wafer, 2 In the third wafer, the ratio of the dies that omit steps S105 and S106 in all dies is 30%, and in the third and subsequent wafers, the dies that omit steps S105 and S106 in all dies. It is also possible to adopt an algorithm that sets the ratio of the above to 40%.
[0094]
Further, FIG. 14 shows a mode in which the transfer proceeds in a spiral manner in the order from the die D at the outer peripheral portion of the wafer 44 to the die D at the central portion of the wafer 44. It can also be adopted. For example, in FIG. 14, after the transfer of one row with the die D1, the die D2, the die D3, and the die D4 is completed, the die D5, the die D30, the die D29, the die D28, the die D27, the die D26 and the next row. It is also possible to adopt a transfer order in which the transfer proceeds to the next transfer and proceeds to one adjacent row thereafter.
[0095]
As mentioned above, although the example of embodiment of the electron beam proximity exposure method which concerns on this invention was demonstrated, this invention is not limited to the example of the said embodiment, Various aspects can be taken.
[0096]
For example, in the electron beam proximity exposure apparatus of the embodiment, the mask mark (M_MX, M_MY, M_MX1, M_MX2, M_MY1, M_MY2) And a second mark mark, a wafer mark (M_WX, M_WY, M_WX1, M_WX2, M_WY1, M_WY2The control of the incident angle of the electron beam is based on the mask distortion of the mask, as in claim 6. A mask mark (M_MX, M_MY, M_MX1, M_MX2, M_MY1, M_MY2) And a second mark mark, a wafer mark (M_WX, M_WY, M_WX1, M_WX2, M_WY1, M_WY2) With respect to the X direction and the Y direction, and the electron beam incident angle control is based on the θ direction deviation between the mask mark and the wafer mark and the mask distortion of the mask. Can be adopted.
[0097]
【The invention's effect】
As described above, according to the present invention, in the die-by-die type electron beam proximity exposure method, the calculated alignment position of the next transferred die and the positioning result calculated from the imaged information are obtained. If the comparison result is equal to or greater than a predetermined value β, alignment is performed so that the comparison result is less than the predetermined value β and / or the incident angle of the electron beam is controlled. On the other hand, when the comparison result is less than the predetermined value β, the transfer is performed as it is. Therefore, the time required for alignment can be shortened as much as possible, and the mask pattern can be easily and accurately transferred to each die of the wafer.
[Brief description of the drawings]
FIG. 1 is a top view of a transfer portion of an electron beam proximity exposure apparatus.
FIG. 2 is a view taken along line AA in FIG.
FIG. 3 is an enlarged top view of a transfer portion of an electron beam proximity exposure apparatus.
FIG. 4 is a diagram used to explain how to obtain the expansion / contraction rate for each chip.
FIG. 5 is an enlarged view of the main part of FIG.
6 is a cross-sectional view taken along line CC of FIG.
FIG. 7 is a schematic configuration diagram of a microscope imaging apparatus applied to the present invention.
FIG. 8 is a diagram used for explaining a method of detecting a positional deviation amount between a mask mark and a wafer mark by a microscope imaging apparatus.
FIG. 9 is a diagram used for explaining a method for detecting a positional deviation amount between a mask mark and a wafer mark by a microscope imaging apparatus;
FIG. 10 is a block diagram showing an embodiment of a control unit of an electron beam proximity exposure apparatus according to the present invention.
FIG. 11 is a diagram showing how the transfer position of the electron beam is shifted by the sub deflector.
FIG. 12 is a flowchart showing an operation procedure of the electron beam proximity exposure method according to the present invention.
FIG. 13 is a diagram used for explaining correction of mask distortion;
FIG. 14 is a diagram showing the arrangement of dies on a wafer.
FIG. 15 is a flowchart for explaining the flow of an electron beam proximity exposure method;
FIG. 16 is a basic configuration diagram of an electron beam proximity exposure apparatus to which the present invention is applied;
FIG. 17 is a diagram used for explaining mask scanning with an electron beam;
FIG. 18 is an overall configuration diagram of an electron beam proximity exposure apparatus to which the present invention is applied.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 15 ... Electron beam 22, 24 ... Main deflector, 26, 28 ... Sub deflector, 32 ... Mask, 44 ... Wafer, 60 ... Electrostatic chuck, 70 ... (theta) XY stage, 90 ... Microscope objective lens, 91 ... Cover glass 91A ... conductive thin film, 92 ... conductive holding member, 93 ... white light source, 97 ... imaging unit, 100 ... central processing unit (CPU), 102 ... signal processing circuit, 104 ... stage drive circuit, 106 ... digital Arithmetic circuit 106A ... Correction arithmetic circuit AX1, AY1, AX2, AY2 ... Microscope imaging device, D ... Die, M_WX, M_WY, M_WX1, M_WX2, M_WY1, M_WY2... wafer mark, M_MX, M_MY, M_MX1, M_MX2, M_MY1, M_MY2... Mask mark

Claims (8)

Proximity of electron beam to transfer a mask pattern formed on the mask to a resist layer on the wafer by placing the mask close to the wafer, aligning each die of the wafer, and scanning the mask with an electron beam In the exposure method,
Step 1 which is a step of storing alignment information of each die in a predetermined process;
Step 2 which is a step of calculating an alignment position of a die to be transferred next from the stored alignment information;
Step 3 which is a step of positioning a die to be transferred next at a transfer position based on the calculation result;
When the die is positioned at the transfer position, the first mark for alignment provided on the mask and the second mark for alignment provided for each die are imaged, Step 4, which is a step of calculating the positioning result of the die from the imaged information;
Step 5 which is a step of comparing the alignment position calculated in Step 2 with the positioning result calculated in Step 4;
Step 6 is a step of performing alignment so that a positional deviation amount between the mask and the die is less than a predetermined value β based on the comparison result and / or controlling an incident angle of the electron beam;
And step 7 which is a step of transferring the mask pattern to a resist layer on the wafer ,
When the comparison result in step 5 is less than the predetermined threshold value α, the frequency that becomes the comparison result is counted, and when the count result reaches the predetermined frequency M, every predetermined number in the subsequent steps. An electron beam proximity exposure method characterized in that step 4, step 5 and step 6 are omitted in the die . 2. The electron beam proximity exposure method according to claim 1 , wherein the transfer of each die of the wafer is performed in the order from the outer peripheral die of the wafer toward the central die of the wafer. Aggregates the comparison result of the step 5 in a single wafer, performs statistical processing, the processing the in the subsequent wafer by the results Step 4, or claim 1 to increase or decrease the frequency of repeated die Step 5 and Step 6 3. The electron beam proximity exposure method according to 2. The alignment in step 6 and / or the control of the incident angle of the electron beam include the amount of displacement in the X direction between the first mark and the second mark, and the Y between the first mark and the second mark. 4. The electron according to claim 1 , wherein the electron is made based on a deviation amount in a direction, a deviation amount in the θ direction between the first mark and the second mark, and a mask distortion of the mask. 5. Beam proximity exposure method. The alignment in the step 6 is performed based on the amount of deviation in the X direction between the first mark and the second mark and the amount of deviation in the Y direction between the first mark and the second mark. 5. The electron beam proximity exposure according to claim 4 , wherein the incident angle control of the electron beam is performed based on a shift amount in the θ direction between the first mark and the second mark and a mask distortion of the mask. Method. The said 2nd mark is provided in multiple numbers around each said die | dye, The said 1st mark is provided corresponding to the said 2nd mark, The any one of Claims 1-5 characterized by the above-mentioned. electron beam proximity exposure method according to any one of claims. 7. The electron beam proximity exposure method according to claim 6 , wherein the second mark is provided on each side of each rectangular die. 8. The electron beam proximity exposure method according to claim 6 , wherein a plurality of the first marks and the second marks are picked up by an image pickup device provided corresponding to the first marks and the second marks. .

Images